Dear Stata Users,
I wanted to ask you more econometric question. Sorry if it is nor relevant for this forum.
Please, help me with the following issue. I want to analyze effect of firm location on its stock preformance (measure monthly) during the crisis. For this I construct the following dummies: Crisis = 1 if (date is within July2007 to March2009 and zero otherwise) Post_crisis = 1 if (date is within April2009 to December2013 and zero otherwise). "Location" is my main variable of interest which is measured at 2006 (year) and is fixed for all the observations. So, my main regression under consideration is as follows:
returni,t=α+β1∗Crisis∗Locationi+β2∗Post_crisis∗Loc ationi+β3∗Crisis+β4∗Post_crisis+β5∗Locationi+γ∗Con trolsi,t+θt+ψi+ϵistreturni,t=α+β1∗Crisis∗Locationi +β2∗Post_crisis∗Locationi+β3∗Crisis+β4∗Post_crisis +β5∗Locationi+γ∗Controlsi,t+θt+ψi+ϵist,
where θtθt - is time fixed effects and ψiψi is firm fixed effects. LocationiLocationi is going to be subsumed because of firm-fixed effects.
My question is: what should be the time period under consideration for running the regression? Shall I run in from 2007 till 2013? Or shall I make it balanced, like from 2000 till 2013? How shall I defend the choice from econometric point of view? Please, help me with this issue.
I wanted to ask you more econometric question. Sorry if it is nor relevant for this forum.
Please, help me with the following issue. I want to analyze effect of firm location on its stock preformance (measure monthly) during the crisis. For this I construct the following dummies: Crisis = 1 if (date is within July2007 to March2009 and zero otherwise) Post_crisis = 1 if (date is within April2009 to December2013 and zero otherwise). "Location" is my main variable of interest which is measured at 2006 (year) and is fixed for all the observations. So, my main regression under consideration is as follows:
returni,t=α+β1∗Crisis∗Locationi+β2∗Post_crisis∗Loc ationi+β3∗Crisis+β4∗Post_crisis+β5∗Locationi+γ∗Con trolsi,t+θt+ψi+ϵistreturni,t=α+β1∗Crisis∗Locationi +β2∗Post_crisis∗Locationi+β3∗Crisis+β4∗Post_crisis +β5∗Locationi+γ∗Controlsi,t+θt+ψi+ϵist,
where θtθt - is time fixed effects and ψiψi is firm fixed effects. LocationiLocationi is going to be subsumed because of firm-fixed effects.
My question is: what should be the time period under consideration for running the regression? Shall I run in from 2007 till 2013? Or shall I make it balanced, like from 2000 till 2013? How shall I defend the choice from econometric point of view? Please, help me with this issue.
